< Back to home

Getting customers to actually rate their orders

Growing a 1% review rate to 15% by removing a barrier nobody had questioned

Company
Magazord, a B2B SaaS e-commerce platform serving 2,100+ merchant clients and millions of end users
Role
Senior Product Designer, owned the work end-to-end
Team
1 Product Manager, and the stakeholders who owned the login requirement I ended up challenging
Tools
Figma, qualitative research, post-purchase journey mapping
The redesigned post-purchase review flow on mobile

Impact at a glance

  • Review submission rate grew from under 1% to 15%, a 15x increase
  • 30% of reviews included user-uploaded images, real social proof for the stores
  • Higher review volume strengthened buyer trust across 2,100+ merchant stores

The context

Fewer than 1% of orders on the platform generated a product review. For a marketplace-style e-commerce network, that is a real problem: reviews are the trust layer that helps the next buyer commit. Across 2,100+ stores, that missing 99% was a quiet, compounding loss of credibility.

Under one percent of orders generated a review

A trust layer that was almost entirely empty

The assumption I tested

The easy read is that customers just do not care enough to review, so the fix is more reminders and more prompts. I did not want to design more nagging on top of a flow I had not examined. So I mapped the full post-purchase journey and looked for where people actually dropped off.

Three friction points came out of it. The biggest one was not a motivation problem at all. To leave a review, a customer had to log in. That single requirement was quietly killing the intent of people who were willing to write something in the moment but were not willing to stop and authenticate to do it.

Post-purchase journey map highlighting three friction points

The largest drop-off was a mandatory login, not a lack of interest

The reframe

Here is where the work stopped being a UI problem. The login wall was not a design decision I could just remove in Figma. It was a business and policy decision that someone above the flow owned, put there for reasons that had made sense at the time.

So I reframed it for the stakeholders. The question was not "how do we make the review form nicer." It was "why are we requiring an account to capture something the customer is already willing to give us." I brought the journey evidence showing where and how much the login was costing us, and made the case that we could keep the integrity we needed with a lighter mechanism. That reframe is what changed the direction: the stakeholders agreed to replace mandatory login with email-only verification.

Reframing the login requirement from a design problem to a business decision

From "improve the form" to "why require an account at all"

The solution

I redesigned the entire post-purchase review flow with a mobile-first approach, since most of the traffic at this point in the journey came from phones. Login was replaced by email-only verification, which kept reviews attributable and trustworthy without forcing the customer to stop and authenticate. The flow was rebuilt to let someone rate, write, and attach a photo in the moment, with as little standing between the intent and the submission as possible.

The redesigned mobile-first review flow with email verification

Results

  • Review submission rate climbed from under 1% to 15%, a 15x increase
  • 30% of reviews arrived with a user-uploaded image, giving future buyers real visual proof
  • The stores gained a trust layer that had been effectively empty before
Review rate grew 15x, from under 1 percent to 15 percent

What I took from it

The 15x number is what people notice, but the move behind it was recognizing that the biggest lever was not inside the design at all. It was a requirement someone else owned, one nobody had questioned. Finding it took research. Removing it took bringing evidence to the people who set the rule and persuading them to change it. The design made the new flow effortless, but the reframe is what unlocked the result.